did you know? the human body only uses 15% of it’s bones at any given time. imagine what would happen if we had access to 50%, or even 80% of our bones
The aristocratic Frankish women – with whom the lords and knights diurnally copulated in the high-ceilinged wooden feasting halls among the packs of dogs and heaped garbage bones of countless red meat roasted dinners – if only to save themselves from constant pregnancies and early deaths in the roulette experiences of perilous childbirths, had begun to urge their masters and sons to fabled and valiant deeds of heroic romance in distant exotic climes.
He shouted: “Eddington! Jeans! Whitebread [an error for Whitehead]! Protons! Neons! These new thingamys – Neutrons – neither this nor that! God, it wearies me! It wearies me!”
In college I did a lot of satisficing. I tried to produce work that was good enough to get a good grade. There was not a lot of reason to do something vastly more impressive than that, and there was always enough work to satisfice at that I never really had time to deeply excel at something.
At work, sometimes I have to just do an adequate job at a lot of work, but that’s not the default state - by default I should have the time and resources to actually do my job well, and to notice and capitalize opportunities to do even better at it. People expect that of me, and they reward me for it, and if I told my boss “I don’t have the time to do anything right because I’m overstretched’ that’d be reason to reassign some stuff.
In college I would have been laughed at if I said “I want to only do one of these essays, but do a really great job at it.”
At work, I often say “I’m going to put in extra hours on project X this week, so I can’t take project Y.”
In college I got stuck a lot. I didn’t know how to do something and I couldn't figure out what my next step was. If I went in to office hours, sometimes I’d get unstuck but sometimes I’d sit in the corner, still unable to articulate the problem. If I didn’t do anything, it’d be a long time before anyone checked in with me - usually after deadlines had been catastrophically missed.
At work if I get stuck, I have a person professionally assigned to get me unstuck. I am professionally assigned to get other people unstuck. There are regular check-ins where the entire topic of conversation is whether I am stuck and on what.
In college you can’t say “I can’t do this because it doesn’t matter.” The point of everything is to evaulate you, so no one cares if it actually changes the world at all.
At work I am expected to prioritize the things that get results. The point isn’t to evaluate me ; the point is to solve a problem.
I learned more in the first six months of a real job than I learned in four-and-a half years of college. And I paid Stanford $220,000 for those years; work paid me.
(Lots of work environments aren’t as nice as mine. But I don’t think this excuses the lack of this kind of thing in colleges. If these are the conditions under which people grow and learn, why does most of higher education still run on inflexible mediocrity-demanding homework assignments with no real support network to make them happen?
Why are we gating the good jobs behind something that is, for many people, much harder than the good jobs, and doesn’t even teach anything about how to do them?)
There was a paper recently where a research team trained a machine learning algorithm (a GAN they called AttnGAN) to generate pictures based on written descriptions. It’s like Visual Chatbot in reverse. When it was just trained to generate pictures of birds, it did pretty well, actually.
(Although the description didn’t specify a beak and so it just… left it out.)
But when they trained the same algorithm on a huge and highly varied dataset, it had a lot more trouble generating a picture to go with that caption. Below, I give the same caption to a version of their algorithm that has been trained to generate everything from sheep to shopping centers. Cris Valenzuela wrapped their trained model in an entertaining demo that attempts to generate a picture for any caption.
This bird is less, um, recognizable. When the GAN has to draw *anything* I ask for, there’s just too much to keep track of - the problem’s too broad, and the algorithm spreads itself too thin. It doesn’t just have trouble with birds. A GAN that’s been trained just on celebrity faces will tend to produce photorealistic portraits. But this one, however…
In fact, it does a horrifying job with humans because it can never quite seem to get the number of orifices correct.
It’s fun to ask it to draw animals though. It knows the texture of giraffes, but not quite exactly their shape. And it knows that boats are on the water, but not necessarily that they are boats.
It also (like many other image recognition algorithms) gets a bit confused about the difference between sheep and the landscapes they’re found on. Other algorithms recognize sheep in pictures of empty green fields. And this one, when asked to draw sheep…
That’s different, though, from asking it to draw *a* sheep. In that case, it knows exactly what to do. It draws the sheep, and then just to be safe it fills the entire planet with wool too.
It really likes drawing stop signs and clocks. Give it the slightest opportunity to draw one, and it will chuck those things all over the place.
Other than its horrifying humans, this algorithm can actually be pretty delightful.
I had way too much fun generating these and ended up with way more than would fit in this one blog post. I’ve compiled a few more of my favorites. Enter your email and I’ll send you them (and if you want, you can get bonus material each time I post).
